171216 01 pogrammheft php k1 - aai.uni-hamburg.de · verb. in a bantu language like kirundi ‘not...

Universität HamburgFakultät für GeisteswissenschaftenAsien-Afrika-InstitutEdmund-Siemers-Allee 1, Flügel Ost20146 Hamburg

asien-afrika-institut

Phasal Polarity in sub-Saharan African languages

International Conference2nd - 4th February 2018

International Conference2nd–4th February 2018

Edmund-Siemers-Allee 1 (Ost)Department of Asia-Africa StudiesUniversity of Hamburg, Germany

keynote speakers: Johan Van der Auwera (Universiteit Antwerpen)Ljuba Veselinova (Stockholms universitet)Dmitry Idiatov (LLACAN - CNRS)

Information under https://www.aai.uni-hamburg.de/afrika/php2018.htmlcontact email address [email protected]


Conference program03.02.2018

8:00-9:00 Registration9:00-11:00 Keynotes Welcome address (Raija Kramer and Roland Kießling) Grammaticalized not-yet markers in Bantu languages (Johan van der Auwera and Ljuba Veselinova) An areal typology of clause-final negation markers in northern sub-Saharan Africa and their historical relation with phasal polarity expressions (Dmitry Idiatov)11:00-11:30 Coffee break11:30-13:00 „Bantu session“ (chair: Tom Güldemann) Nondums in Bantu languages: types, variation and paths of change (Maud Devos) Phasal Polarity in the Nguni cluster (Bantu, S.40) (Axel Fleisch) The expression of Phasal Polarity in Cuwabo (Bantu P34) (Rozenn Guérois)13:00-15:00 Lunch15:00-16:30 „Grammaticalization session“ (chair: Gabriele Sommer) Phasal Polarity in Languages of the Wider Lake Chad Area – From “First” to “Yet” (Georg Ziegelmeyer) Phasal Polarity expressions in Fulfulde of northern Cameroon: Constructional variation and grammaticalization (Raija Kramer) From Phasal Polarity expression to aspectual marker: Grammaticalization of already in Colloquial Singapore English (Lijun Li and Peter Siemund)16:30-17:00 Coffee break17:00-19:00 „West Africa session“ (chair: Raimund Kastenholz) Phasal Polarity in Ewe: Diversity of constructions and dialect differences (Felix Ameka) Phasal Polarity particles in Kusaal (Hasiyatu Abubakari) Phasal Polarity in Ngiemboon (Eliane Kamdem) Phasal Polarity expressions in Bambara (Mande) (Klaudia Dombrowsky-Hahn)


17:00-19:00 „West Africa session“ (chair: Raimund Kastenholz) Phasal Polarity in Ewe: Diversity of constructions and dialect differences (Felix Ameka) Phasal Polarity particles in Kusaal (Hasiyatu Abubakari) Phasal Polarity in Ngiemboon (Eliane Kamdem) Phasal Polarity expressions in Bambara (Mande) (Klaudia Dombrowsky-Hahn)19:30- open Conference diner Music: DJ Hauke Dorsch (African Music Archive, JGU Mainz) Pamoja-Chor Hamburg

04.02.2018

10:00-12:00 „East Africa (non-Bantu) session“ (chair: Rainer Vossen) Phasal Polarity in Omotic: Is it the verb again? (Bernhard Köhler) Negation and the expression of Phasal Polarity in Kambaata (Cushitic) (Yvonne Treis) Phasal Polarity items in Datooga (Nilotic): Interactions with tense, aspect, and participant expectation (Alice Mitchell) Phasal polarity in Iraqw and other Tanzanian languages (Maarten Mous)12:00-13:00 Lunch13:00-15:00 „Varia session“ (chair: Ines Fiedler) Phasal Polarity expressions in Tashlhiyt (Christian J. Rapold) Phasal Polarity in Isu – and beyond (Roland Kießling) Phasal Polarity in Suma (Gbaya language, northern CAR) (Raymond Boyd) Phasal Negation in Zande (Helma Pasch)15:00-15:30 Farewell

For your meals during the breaks you can thrive off the diff erent culinary off eringssurrounding the university. In the map we listed our favourite places to eat.

Balutschi

Spiesserei

Abaton

Gran Sasso

BrodersenRoxie

Abaton: bistro serving provencal french food in a homely atmosphere Grindelhof 14A open 12am - 12pm

Brodersen: restaurant with nautical fl air off ering fi sh specialties from Hamburg Rothenbaumchausee 46 open: 5pm - 11pm

Roxie: Raija‘s and Roland‘s favourite place serving brunch, cocktails and international dishes Rentzelstraße 6 open: 11am - 2am

Balutschi: Pakistani and Indian kitchen, good vegetarian and vegan dishes Grindelallee 31 open: 12am - 11pm

Gran Sasso: italian restaurant off ering pasta, pizza and fi sh Schlüterstraße 12 open: sat 11:30am - 11pm sun: 5:00pm - 10pm Spiesserei: modern restaurant specialised in tapas and grilled dishes Grindelallee 32 open: 11:30am - 10:30pm closed on sundays

Abstracts


Phasal Polarity in Ewe:Diversity of constructions and dialect differencesFelix K. Ameka, Leiden University Centre for Linguistics

In this presentation, I show that the expressions for the phasal polarity concepts related to persistence or non-persistence of a situation (STILL, NO LONGER) and the realization or not of an expected situation (ALREADY, NOT YET) do not constitute a grammatical paradigm in Ewe. Thus, Ewe is different from several European langu-ages where these notions are encoded in terms in one grammatical class (cf. Van der Auwera 1998). STILL is expressed by a periphrastic construction involving the repetitive preverb marker ga ‘again’ reinforced by the intensifier ko ‘only, just’. This construction can also be combined with an alternative continuative phasal aspect construction involving the suppletive locative verb le ~ nɔ ‘be.at:PRES~NPRES’ and a complement headed by the postposition dzi ‘upper surface’. NO LONGER is also expressed using the repetitive preverb ga ‘again’ but, in this case, together with the clausal negator. ALREADY is expressed by an adverbial xoxo ‘already’ while NOT YET is coded in negative clauses by either the preverb auxiliary kpɔ́ ‘COUNTEREXPECTATION’, grammaticalised from the verb kpɔ́ ‘see’ (cf. Heine et al. 1991), or the negative pola-rity adverbial haɖe(ke) ‘yet’. Both NOT YET constructions can be combined and they also manifest dialectal variation. The former is predominant in Southern dialects like Anlo and Tɔŋu, the latter is used in the Inland dialects. The ways in which these structures interact with phasal aspect, imperfective and habitual constructions as well as focussing intensifiers will be discussed.

Phasal Polarity in SumaRaymond Boyd, LLACAN-CNRS

The asymmetry of the ALREADY/NO LONGER pair with respect to STILL/NOT YET is dis-cussed. The polysemy of terms used for ALREADY, which varies widely with the indi-vidual language, is remarked, and the example of Suma (a northern Gbaya language of the CAR) is adduced. In this language, the particle covering the ALREADY sense for which NO LONGER is proposed as polar opposite has the primary senses of ‘past no longer relevant to the present’ (in French, the révolu) and the classic completive sen-se ‘firmly intended, about to occur’ as when, for example, the verb ‘go’ in the com-pletive form is used to mean ‘(x) has decided to go, is about to leave’. The NO LONGER sense is obtained by use of the particle ‘again’ in a negative utterance.STILL/NOT YET is expressed by a verb meaning ‘remain’ and its negation used for pro-visional situations and contrasting with a verb ‘sit; be in a final state’ remindful to some extent of Spanish estar vs. ser. The very broad semantic range of this verb is presented in detail, one of its striking features being the possibility to express both a lasting state and the nearly enantiosemic ‘as soon as, immediately thereafter’. A sequence of senses giving rise to this situation is described.

Nondums in Bantu languages:Types, variation and paths of changeMaud Devos, Royal Museum for Central-Africa

Bantu languages are known to be verb centered (Nurse 2008: 21). Many categories including TAM but also polarity, focus, deixis, taxis, etc. are typically expressed on the verb. In a Bantu language like Kirundi ‘not yet’ is rendered entirely by verbal affixes, i.e., a combination of the standard negative prefix and the inceptive prefix occurring before and after the subject prefix, respectively. In Kirundi negative ‘not yet’ is symmetrical with affirmative ‘already’, as seen in (1). However, in many Bantu languages ‘not yet’ constructions involve isolated negati-ves which cannot be formally linked to an affirmative form. Based on a convenience sample of 100 languages this paper investigates ways of expressing ‘not yet’ in Ban-tu. At least three types can be distinguished. The first involves the use of a negative phasal adverb, as seen in (2). The second makes use of a persistive auxiliary void of negative morphology, illustrated in (3). The third combines standard negative mar-king with a (dedicated) tense marker, as in (4), a (dummy) auxiliary, as in (5), or both, as in (6). If symmetrical negation is involved, we investigate whether nondums al-ways pair with iamitives, the Kirundi case, or whether other pairings are attested.After identifying these and possibly other types, we zoom in and make hypotheses about possible paths of change through attested micro-variations in the expression of ‘not yet’.

(1) Kirundi JD62 (Meeussen 1959 cited in Schadeberg 1990: 7-9)a) ‘already’ b) ‘not yet’

tu-ráa-kúbuur-a ńti-tu-ráa-kúbuur-aSP1PL-INC-turn-FV NEG-SP1PL-INC-turn-FV‘Have we already turned’ ‘We are not turning yet’

(2) Luvale K14 (Horton 1949: 162)a) kanda vamanyise kulya

NOT_YET SP2-finish-SBJV 15-eat-INF‘They have not yet finished eating’

(3) Manda N11 (Bernander 2017: 263)a) a-akona ku-lemb-a

SP1-PERS 15-write-INF‘she hasn’t written yet’ (litt.: ‘she is still to write’)

(4) Matuumbi P13 (Odden 1996: 66)ni-ná-kalaang-a li̧SP1SG-NOT.YET-fry-FV NEG

‘I haven’t yet fried’


(5) Shona S10 (Fortune 1955: 353)kwete, ha-ta-ti ta-i-on-ano NEG-SP1PL-PST-AUX SP1PL-PST-OP9-see-PART

‘No, we have not yet seen it.’

(6) Shangaji P312 (fieldnotes)si-ná-tthí o-cuúw-aNEG.SP1SG-INC-AUX 15-know-INF

‘I did not yet know’

Phasal Polarity expressions in Bambara (Mande)Klaudia Dombrowsky-Hahn, Universität Bayreuth

In Bambara there are at least two expressions for each of the four PhP concepts:

(a) kàban ~ kà bán, kàkɔrɔ and kélen express the ALREADY concept (b) bán, fɔ́lɔ express the NOT YET concept (c) túgun and bìlen~blèn express the NO LONGER concept (d) the borrowed hálisa [hálísà], [hálìsà], and, rarely, bìlen~blèn express the STILL concept.

The expressions under (a) and (d) are found in affirmative clauses, displaying one of the affirmative predicate markers (auxiliaries or copulas), (b) and (c) in clauses bea-ring a negative predicate marker.At least some items occur in more than one phasal category, justifying the label ‘fle-xible’ for the system. kàban ‘already’ and bán ‘not yet’ show the common element bán, whose origin, the verb bán ‘finish, end’ is transparent. When the borrowed expression hálisa and the rare – and maybe only recently de-veloping bìlen expressing the concept ‘still’ are disregarded, the phasal system of Bambara can be qualified as asymmetric, comprising the three periods 1) not yet; 2) already; 3) no longer.

period 1 period 2 period 3NOT YET ALREADY NO LONGERbán, fɔ́lɔ kàban~kà bán túgun,bìlen~blèn

+ - +

Asymmetric phasal system of Bambara (borrowed hálisa is excluded)The expressions of the ALREADY and the NOT YET concepts differ with respect to prag-matic sensitivity. Among the first, kàban and kàkɔrɔ are used in neutral scenarios, whereas kélen is the expression found in counterfactual contexts. Interestingly, the expressions hálísà or háli bì, the first part of which is borrowed from Arabic and that express in Bambara the occurrence of the situation implying a reference point in the

subsequent phase, is also used as a strongly marked NOT YET expression ‘still not’, emphasising the extendedness of the phase of non-occurrence before the expec-tation of its overdue occurrence. However, these items do not belong exactly to the same paradigm as the other phasal polarity items.

Phasal Polarity in the Nguni cluster (Bantu, S.40)Axel Fleisch, University of Helsinki

In South African Nguni varieties, phasal polarity notions are expressed by way of af-fixes in the fairly complex verbal morphology. The concepts STILL and NO LONGER are formally related: The verb form containing the persistive affix -sa- ‘still’ (1) yields the “no longer”-construction when externally negated (2).

(1) si-sa-hlal-a e-Mosesetjana1PL-still-live-FV LOC-Mosesetjana‘We still live in Mosesetjana’

(2) a-si-sa-hlal-i e-MokopaneNEG-1PL-still-live-FV.NEG LOC-Mokopane‘We are no longer living in Mokopane’

The notion ALREADY is expressed with a pre-verb sele (with a shorter variant, the prefix se-).

(3)a sele ba-fik-ile ekhaya (3)b se-si-fik-ilealready 3PL-arrive-PFV at.home already-1PL-arrive-PFV

‘they have already arrived at home’‘we have already arrived; ´we’re here now’

Besides describing basic Nguni PhP mechanics, the contribution discusses evidence for neutral v. counter-expectational usages of sele/se- and persistive sa-, telicity (se- ‘already’ being rather general), and strategies to paraphrase the NOT YET gap. This follows the lead of the position paper on PhP (Kramer 2017). Lexical verbs in Ngu-ni interact with grammatical tense-aspect in intricate ways (Crane & Fleisch 2016, forthcoming; cf. also Croft 2012, Bar-el 2015). Phasal notions as expressed by sele/se- and sa- are obviously sensititive to these properties; this paper documents relevant observations and constraints discovered in systematic elicitation. Interestingly, a substantial degree of microvariation with regard to these parameters exists in what seems otherwise a fairly homogenous system across the Nguni varieties.


The expression of Phasal Polarity in Cuwabo (Bantu P34)Rozenn Guérois, SOAS University of London

Expressing the four concepts lying behind the notion of phasal polarity (PhP), i.e. ‘al-ready’, ‘still’, ‘not yet’, and ‘no longer’, in Bantu languages usually involves the use of different strategies. This talk particularly focuses on PhP in Cuwabo (Mozambique, Bantu P34), which presents the following characteristics:

Table: Phasal polarity in CuwaboPhP concept ‘already’ ‘still’ ‘not yet’ ‘no longer’Form -hi- / -ile (+ ja) =vi -ná- (+ NEG) viína (+ NEG)Status Verbal affix

(+ Adverb)Verbal enclitic Verbal prefix Adverb

Function(s) Perfective (+optional adverb ja ‘already’)

Restrictive Counter-Expectational

‘too’ > ‘no longer’with negative verbs

Example (já) o-hí-dhówá‘he went already’

ónówííbávi‘he is still singing’

ka-ná-kwa‘he has notdied yet’

kaneéddíle víina‘he no longer worked’

In this talk, I will describe into details and illustrate the coding of PhP expressions in Cuwabo, and seek to classify them along the six parameters developed by Kramer in her position paper.

Phasal Polarity particles in KusaalHasiyatu Abubakari, University of ViennaOne of the least addressed topics in a less researched language is PhP expressions in Kusaal, a Mabia (Gur) language spoken in Ghana, though this language interestingly has various particles for expressing almost all the four identified PhP expressions in English (Van Baar 1997).

1. kʋn, nan ‘already’ 2. kʋn, kpɛŋ ‘still’ 3. nan ‘yet’ 4. nan pʋ, kʋn pʋ ‘not yet 5. pʋ lɛŋ ‘no longer’

Kusaal can be said to have a flexible PhP coverage with identical particles used in ex-pressing more than one PhP notion. The questions to be addressed include: (i) what are the various structural and semantic parameters of PhP particles in Kusaal? And (ii), are there any constraints underlying the use of PhP particles in the language? These will be addressed following Kramer (n.d.).

An areal typology of clause-final negation markers in northern sub-Saharan Africa and their historical relation with phasal polarity expressionsDmitry Idiatov, LLCAN-CNRS

Clause-final negation markers (CFNMs), although typologically rare, can be found in a very wide range of languages of northern sub-Saharan Africa (NSSA). Given that on the one hand, CFNMs tend to occupy the same constructional slot as phasal polarity expressions (PPEs), and on the other hand, strong semantic and often also formal links are known to exist between PPEs and negation, CFNMs may be expected to often develop out of PPEs. However, this expectation is not borne out by the data available on the development of CFNMs in the languages of NSSA.This paper has two objectives. First, based on a sample of 618 African languages, it provides an analysis of spatio-temporal language dynamics in sub-Saharan Africa with respect to the feature CFNM. I argue that it is important to consider together the languages that have the feature under investigation and the languages that do not have it. Furthermore, in order to better capture the diversity of the languages that have CFNMs, I increase the degree of granularity of my data by taking into ac-count two parameters, viz. obligatoriness of CFNMs and possible restrictions on the freedom to use CFNMs in different constructions. The spatial analysis of the data suggests that we need to distinguish two focal areas of the feature CFNM. The first one, the Central Focal Area, is the most prominent of the two and spans the east of West Africa and parts of Central Africa. The second one, the Western Focal Area, is less prominent and is restricted to West Africa. The two focal areas are separated by a major discontinuity around Ghana, Togo and Benin.Second, I discuss a number of cases from Mande languages where CFNMs happen to be historically related to PPEs. In line with the general rarity of the change from a PPE to a CFNM, CFNMs originating in PPEs tend to remain restricted to certain TAM constructions and not to evolve into default CFNMs. In those rare cases where PPEs do evolve into default CFNMs, this evolution is not direct and necessarily proceeds through the addition of an indefinite semantic component (such as ‘not yet’ > ‘not ever yet, not on any occasion yet’) and the development of the implicature of an intersubjective operator processing hearer-sided expectations and presuppositions (such as ‘not ever yet, not on any occasion yet’ > ‘not at all, really not’).

Phasal Polarity in NgiemboonEliane Kamdem, University of Bayreuth

The Ngiemboon (Bamileke, Grassfields Bantu) PhP paradigm consists of four items: ńdáʔ/láʔ “already”, ɲé “still”, ʧʉ́ʔ…wɔ́ “no longer” and ɲé tə̀…wɔ́ “not yet”. In this talk, I will provide information about the semantic as well as formal features characteri-zing these PhP items. The discussions will include elements such as the relatedness in form of two of the four PhP items (ɲé “still” and ɲé tə̀…wɔ́ “not yet”), the coding of the PhP items alike in both neutral and counter-expectation scenarios, the use of the “already” ńdáʔ/láʔ and “no longer” ʧʉ́ʔ…wɔ́ items to express early turning points, the derivation of PhP items/auxiliaries from verbs (e.g.: ńdáʔ/láʔ “already” origina-tes from the verb ńdáʔ/láʔ “to spend the night”) and ability of these auxiliaries to


keep some of their verbal properties and the characterization of the PhP paradigm in Ngiemboon as internally symmetric and externally partially symmetric.

Phasal polarity in Isu – and beyondRoland Kießling, University of Hamburg

The phasal polarity system in Isu, a West Ring language of Grassfields Bantu (Ca-meroon), seems to be restricted to a single item, i.e. nám(ə́) which is fully dedicated semantically to encode the concept STILL. Other phasal polarity notions are partly covered by polysemous items which have their semantic core elsewhere, e.g. the concept NO LONGER is expressed by the adverbial kám(ə́) ‘again’ in externally nega-ted constructions. The English loan ɔ̀lrɛ́dì is frequently used to fill the ALREADY gap, otherwise partially covered by the adverbial tə́ŋ(ə́) ‘readily, promptly’.From a formal point of view, all items mentioned belong to the category of hybrid adverbials, a type of defective verbs which have become grammaticalised in their position of minor verbs or co-verbs in asymmetrical serial verb constructions. They betray their verbal origin by virtue of their participation in aspectual morphology, while they are synchronically rather defective with respect to verbhood, since they cannot establish single-handedly a verbal predication. Across the Ring languages there is considerable variation with respect to the expres-sion of phasal polarity, since nearby Central Ring languages such as Men come up with a fundamentally different system, providing dedicated items for the concepts ALREADY (kàinə̀) and NOT YET (ɲə̀m + NEG), while the concept STILL is expressed by the adverbial pá which is not cognate to WR *nám(ə́).The contribution gives an overview of phasal polarity in Isu generally, but will have its limitations concerning semantic detail due to the fact that it relies on mining a limited corpus of some thirty texts of different lengths representing various genres, without being supported by any elicitations geared towards clarifying phasal pola-rity concepts.

Phasal Polarity in Omotic - Is it the verb again?Bernhard Köhler, Goethe-Universität Frankfurt am Main

The Omotic languages of southwest Ethiopia are well-known for their complex ver-bal morphologies, which can transform a CVC- verb root into a verb form of up to five syllables. Therefore, it is tempting to assume that phasal polarity is marked by verbal morphemes in these languages. Some constructions such as the so-called completi-ve in Wolaitta, which may be used to express ‘already’ (Wakasa 2008: 963), seem to suggest this. Other constructions, however, contain full-fledged adverbs to express phasal polarity: in Oyda, for example, ‘still’ is rendered by a separate word obviously derived from ‘now’ (own field notes). The present paper will explore phasal polarity in Omotic languages from a typological perspective, using examples from different languages. Not only the syntactic status of the markers, but also their origins and possible other functions will play a role in the discussion.

Phasal Polarity expressions in Fulfulde of northern CameroonConstructional variation and grammaticalizationRaija Kramer, University of Hamburg

Fulfulde is the lingua franca in northern Cameroon and, by now, it is L1 not only of “ethnic” Fulani, but also of speakers outside this group. Based on a corpus of biblical texts and (elicited) data collected among 15 speakers with different linguistic/ethnic backgrounds, I will show means of expressing phasal polarity in this Fulfulde variety. Negative PhP categories NOT YET and NO LONGER are homogenously expressed by constructions involving a negator and the clause final elements fahin ‘no longer’ and tawon ‘not yet’. However, constructional variation (aside from numerous periphra-stic ways of rendering PhP meaning) can be found in the coding strategies of posi-tive PhP categories ALREADY and STILL. ALREADY is left unmarked or denoted by clau-se-final elements timmi/jinni derived from ‚finish‘-verbs timm-/jinn-. STILL is overtly unmarked or coded by either the NOT YET item tawon or the NO LONGER item fahin. In my presentation, I will demonstrate that the occurring variation in ALREADY/STILL expressions depends on the type of PhP-scenario to be coded (NEUTRAL or COUNTER-FACTUAL), and on the linguistic backgrounds of the speakers possibly responsible for rather recent contact induced grammaticalizations of PhP constructions.

From Phasal Polarity expression to aspectual markerGrammaticalization of already in Colloquial Singapore EnglishLijun Li and Peter Siemund, University of Hamburg

Our paper focuses on the aspectual marker already in Colloquial Singapore English (CSE) and argues that the CSE already has changed its grammatical function from what is referred to as a Phasal Polarity (PhP) expression implying a reference point at a prior phase in English as well as in other European languages (cf. Van der Auwera 1993; Van Baar 1997) to an aspectual marker marking the perfective aspect (both the completive and inchoative aspect), as shown in (1) and (3). We submit that such grammaticalization results from contact induced interference introduced by the ubi-quitous Sinitic substrates (i.e. Hokkien, Cantonese, and Mandarin) spoken in the area (see (2) and (4). Our research draws on material from the Oral History Interviews held by the National Archives of Singapore as well as the Singaporean component of the International Corpus of English (ICE-SG), which provides evidence that higher ratios of already are indicative of higher substrate influence. We will extend our research to the additive marker also, the experiential marker ever and the contrastive foci maker one, three other expressions in CSE which have acquired Chinese-derived grammati-cal functions.

(1) Colloquial Singapore English (Completive)a. I see the movie already. (Bao 2005: 239)

‘I saw the movie.’ or ‘I have seen the movie.b. I stopped already working. (Oral History Interviews 000284/Reel 5)

‘I have stopped working.’


(2) Mandarin ChineseZuótiān wǒ kàn le zhè ge diànyǒng.Yesterday I see le this CL film.‘I saw the film yesterday’ (CLL Comtemporary/Prose 3/Xiao Q.)

(3) Colloquial Singapore EnglishHe already studied there. But after some time, I went there, I can‘t find him. (Oral History Interviews 000009/Reel 23)‘He had started to study there, but I didn’t find him when I went there’ (Inchoative)

(4) Mandarin Chinesewǒ érzi shàngxué lemy son go school ASP‘My son has started school.’

Phasal Polarity items in Datooga (Nilotic)Interactions with tense, aspect, and participant expectationAlice Mitchell, University of Bristol

In Datooga, a Southern Nilotic language of Tanzania, the PhP concepts STILL and NO LONGER are encoded by means of a continuative verbal prefix ud(u)-, e.g.:

(1) g-úd-dá-gáw-ìi-s-chìAFF-CONT-1SG.SBJV-milk-PLUR-TERM-AP‘I’m still milking’

The concept ALREADY is encoded in one Datooga dialect by means of an auxiliary verb, but appears only rarely in my corpus of conversational Datooga. There is no specialized item for NOT YET, a concept typically expressed by means of the negative perfect. Such a gap in the PhP system is not predicted by Van Baar’s (1997) “expres-sibility hypothesis”.This paper first looks at formal characteristics of PhP items in Datooga and then explores the semantics of the continuative udu- prefix, showing that its meaning interacts with tense, polarity, and, importantly, lexical aspect. While verbs denoting states and activities combine with ud(u)- to give straightforward STILL or NO LONGER meanings, verbs denoting achievements and accomplishments behave differently. Of particular note is the use of the negated continuative prefix with achievement verbs, which indicates that a state of affairs was expected to take place, but did not (and will not) occur. This PhP expression, which exclusively codes counterfactual scenarios, presents a configuration of phasal polarity that has received minimal at-tention in the literature.

Phasal polarity in Iraqw and other Tanzanian languagesMaarten Mous, Leiden University Centre for Lingustics

The paper investigates how phasal polarity is expressed in Iraqw (Cushitic, Tanzania). All shades of phasal polarity are expressed in Iraqw by a number of markers. The closest to designated markers for phasal polarity are two verbal adverbs malé ‘again’ and qaro ‘already’. However the concepts of ‘already’and ‘again’ can also be expres-sed by other ways including tense/aspect markings on the verb which have wider functions. The paper provides an overview of the adverbs malé and qaro, their se-mantics, pragmatics and behaviour under negation. I position these two adverbs in the system of verbal adverbs in Iraqw and conclude with some questions about the relevance of the type of morphological expression for the function of phasal polarity items by contrasting Iraqw to some Tanzanian Bantu languages that have phasal polarity “tenses”.

Phasal Negation in ZandeHelma Pasch, University of Cologne

Standard negation in Zande is bipartite, the first negator nga, probably derived from the homo-phonous copula, directly following the verb, the second, te, identical with the negator of Lingala, being in clause-final position (1). In subordinate clauses and commands, the second negator is ya (2).

(1) Ani a-kpi nga nyamu ko te …1p.1 III-covet NEG1 love 3m NEG2We do not like him …

(2) Mo di parapluie tipa ka mu a-ziro nga ya.2s.1 take umbrella in.order SUB 2s.1 III-get.wet NEG1 NEG2Take an umbrella so that you won’t get wet.

Eliciting examples of phasal negation of the not yet and no longer type causes pro-blems just like eliciting examples of the already and the still type. And in texts there are no really clear examples. Many examples may also be imderstood in a non-pha-sal reading and speakers are reluctant to give equivalents of single short clauses. In most examples phasal notion is evident only from context.Already is indicated by ta (< taa ‚stretch, extend‘) (3) or ima (< ima ‚dwell, remain‘) (4) preceding the verb. Note that ta which also has the notion finally (5) (cf. van der Auwera 1993: 618ff) functions quite generally as a marker of similitude. The specific notion reveals from context. While ta is found with all verbs, including those descri-bing states and change of states, ima is found with only with action verbs.

(3) rago ki ta gira …place SEQ yet shinewhen it was already morning …


(4) bakure-mi nga Ngbia ima fu a-ge fe-re.blood.brother-1s.1 cop Ngbia already give pl-termite for-me‘My friend Ngbia has given me termites.

(5) Gbara ki fu a-fu kindii, si ki ta wi,yam SEQ stew III-stew continually INAN.1 SEQ yet cookedWhen the yam had boiled and was cooked

Not yet is likewise indicated by ta, now in combination with the negators nga and te (6,7).

(6) ko a-ta ye nga te3m III- come NEG NEGHe has not yet come

(7) Gomoro na-manga re kindi, mbiko mi ta ri nga he tehunger II-make 1s.2 continually because 1s.1 eat NEG thing NEGI am still hungry, because I have not eaten yet.

When opposed to not yet, already is always expressed by ima, not by ta (8).

(8) Petro ima ndu na gara ue ku dumo waraga yo,go year two DIR house learning there

Peter has been (going) to school now for two years

ono ku ta ino nga ka ke pai tebut 3m know NEG SUB write thing NEGbut he does not yet know, how to write

No longer is indicated by the adverb berewe, which also has the notion ‚again‘. It follows the predicate but precedes the negator in clause final position.

(9) Ka Petro ngbisi nga gu bus re berewe te.SUB P meet.PERF NEG DEF.D bus DEM again NEGPeter will not catch the bus.

(10) Wa mi a-du na a-gude bisue boti sa,like 1s.1 III-be PL-child five on.top oneWhen I had six children

mi a-ida nga ka batika berewe te.1s.1 III-want NEG SUB beget again NEGI did not want to have more children.

Still is likewise indicated by the adverb kindi(i) (11, 12) the basic notion of which is ‚longtime, continually‘. In combination with action verbs and in complex structures containing several verbs, it is expressed by ta (13.)

(11) Gomoro na-manga re kindihunger II-make 1s.2 longtimeI am still hungry …

(12) rago na-bira kindiiplace II-black continuallyit is still dark

(13) ta mbira-ha mi ga a-gadrink-INAN.2 1s.1 go.home III-go.home

I went home [e.g. driven out of a bar] while still drinking

Phasal Polarity expressions in TashlhiytChristian J. Rapold, University of Regensburg

The primary means of expressing phasal polarity in Tashlhiyt, a Berber language of Morocco, are the three particles yad, sul and ta:

yad ur sul ur ta sulalready NEG still NEG yet still‘already’ ‘no longer’ ‘not yet’ ‘still’

The morpheme ur ‘NEG’ is the standard negation marker in predications. Yad ‘alrea-dy’ and ta ‘yet’ are dedicated PhP markers, whereas sul ‘still’ is mildly polysemous. Sul ‘still’ is grammaticalized from the verb sul ‘to persist, still be’, yad likely has a nominal origin. The notion ‘ALREADY’ is overtly expressed much less frequently than in English, French or German, which places Tashlhiyt in the vicinity of languages that have a structural gap here (Van Baar 1997, van der Auwera 1998). In a corpus of English, French and German translations of Tashlhiyt texts, only about 5% of the occurrences of already, déjà and schonPhP have a corresponding PhP particle (yad) in the Tashl-hiyt original.Form and meaning of these PhP markers as well as the constructions they enter into will be described in detail following the typological parameters in Kramer (2017), mainly based on data from published texts.


Negation and the expression of Phasal Polarityin Kambaata (Cushitic)Yvonne Treis, LLACAN

Kambaata, a Cushitic language of Ethiopia, has five suffixes for the negation of in-dicative main verbs, imperatives, jussives converbs and relative verbs. The language has not grammaticalized the expression of phasal polarity. The first part of the pa-per gives a brief overview of the negative system of the language and discusses the function of the five inflectional negative morphemes as well as of derivational nega-tive morphology and negative pronouns. The second part is dedicated to a recently grammaticalized morpheme kása, which only occurs in interrogative and negative contexts. It expresses that something was never done up to now or inquires whether something was ever done before. Based on a corpus of biblical texts and the Little Prince, I will investigate then in the third part of my paper how phasal polarity mor-phemes of the translated language are rendered in the Kambaata translation. We will see that ‘already’ and ‘not yet’ is usually ignored by the translators, whereas ‘still’ and ‘no longer’ are periphrastically expressed.

Grammaticalized not-yet markers in Bantu languagesJohan van der Auwera, Universiteit AntwerpenLjuba Veselinova,Stockholms universitet

Bantu languages are typically used to illustrate the distinction between plain negati-on and the encoding of both the non-realization of a situation, and the expectations of the speaker for its future realization. Data from Digo, a Bantu language from Ke-nya and Tanzania, are shown in (1) by way of introductory illustration.

(1) Digo [digo1243], (Nicolle 2013: 150, 135)a. u-ka-rim-a b. ta-m-ka-fwih-a

2SG-ANTERIOT-farm-FINAL VOWEL NEG-2PL-ANTERIOR-dance-FINAL VOWEL‘You have farmed/you farmed’ ‘You have not danced’

c. ta-ri-dzangbwe-dung-aNEG-5-INCEPTIVE-pierce-FINAL VOWEL‘It has not yet pierced’

In Digo negation is expressed by the prefix ta- for a number of tenses, including the anterior, cf. (1b). The form –dzangbwe- is glossed ‘INCEPTIVE’; it is a bound item ob-served only in the context of negation or in questions. That is, it is an example of a negative polarity item, hereafter NPI. The suffix –dzangbwe- indicates “that an event has not occurred prior to and including the reference time” (Nicolle 2013: 157). This author goes on to state that this form typically also implies that the non-realized event may occur at some point after the reference time. The form –dzangbwe- appe-ars in the same position as other tense-aspect markers in Digo; thus it can be seen to be in opposition with them. It is considered a grammatical rather than a lexical item based on the following features (i) it is bound, rather than free form; (ii) it has a fixed position within verb forms, (iii) it is restricted to specific contexts, e.g. negation and questions, and (iv) it has an abstract rather than specific meaning. Following the praxis suggested by Bybee, Perkins and Pagliuca (1994: 2), grammatical markers are also referred to as grams. For discussions of similar expressions, see Comrie (1985), Contini-Morava (1989), van Baar (1997), van der Auwera (1998), Plungian (2000), Zes-han (2004), the list of references given here is minimal. Olsson (2013) as well as Dahl & Wälchi (2016) discuss the categories of perfect and iamitive from a cross-linguistic perspective. The latter categories are relevant here since not-yet markers are often cited as special negators for them. As indicated above, it is often stated that markers such as –dzangbwe- are common-ly encountered in Bantu languages. However, there is still no specific information about their frequency in the Bantu family. Consequently, the first goal of this study is provide a plausible picture of the distribution of grammaticalized not-yet expressi-ons in this family. We also strive to outline their contexts of use and ultimately arrive at a semantic map of this under-described lexico-grammatical category. We work with a sample of 100 Bantu languages fairly well distributed throughout the Bantu area. Our main data sources are grammars and parallel texts, specifically the Parallel Bible Corpus. We also use a questionnaire for data elicitation but this is only feasible for a handful of languages. So far the preliminary results indicate that not-yet markers are abundant in the cen-tral-eastern parts of the Bantu territory but are not so common in the north-west areas. It is not always easy to determine the morpho-syntactic category of not-yet markers. In some languages, such as Digo above, they are clearly affixes, typically, prefixes. However, in many others, not-yet markers appears as defective verbs or auxiliaries. There are also instances of more complex constructions. In terms of cont-ent, not-yet markers typically indicate the non-occurrence of an otherwise expected action or state but also an anticipation about its occurrence. They are also frequently used as temporal subordinators. Other uses observed in the parallel texts are indica-tion of surprise/counter-expectation, emphatic negation, questions and near future. There are also a few instances where the not-yet marker appears to be used as a marker of plain negation.


Phasal Polarity in Languages of the Wider Lake Chad AreaFrom “First” to “Yet”Georg Ziegelmeyer, University of Vienna

During the past decade the study of geographical typology and linguistic areas at-tracted more and more attention in African Studies. On the one hand some scholars have claimed that Africa as a whole can be considered a linguistic area, i.e. Africa can be defined as a linguistic area vis-à-vis the rest of the world. On the other hand se-veral scholars reject the idea of Africa as a single linguistic area, but identify smaller linguistic areas on the continent. Against this background we coined a linguistic micro-area which we labelled the Wider Lake Chad Area (WLCA), and which is characterized by mutual contact pheno-mena across genealogically unrelated languages. Besides borrowing of content and function words, convergence of typological structures constitutes a major clue for the justification of our linguistic area at issue.When it comes to the expression of phasal polarity in the languages of the WLCA establishing clear genealogical patterns across Chadic, Saharan, and other langua-ges of the region appears to be difficult. On the one hand insufficient information on phasal polarity in the various grammars and dictionaries do not allow us to draw hard-and-fast conclusions. On the other hand certain strategies for the expression of phasal polarity crisscross genetic boundaries, e.g. Kanuri as well as several Chadic languages may use a verb meaning “precede, be first at, arrive before, etc.” in order to express the notion of “have already done”. Last but not least, it turns out that several languages of the WLCA borrowed the Kanuri adverbial adjunct duwô ‘first, (not) yet, etc.’ for the expression of concepts like “still, first, (not) yet”.In my presentation I will give a brief overview of the various strategies found in the languages of the WLCA to express phasal polarity, before I concentrate on gram-maticalization processes of Kanuri duwô ‘first’ into various grammatical domains. Suffice it to say at the moment that over the past hundred years probably no other grammatical marker attracted more attention in Kanuri studies than the so called adverbial adjunct duwô.

171216 01 pogrammheft php k1 - aai.uni-hamburg.de · verb. in a bantu language like kirundi ‘not...

Documents